MCPcopy Create free account

hub / github.com/MAC-VO/MAC-VO / functions

Functions937 in github.com/MAC-VO/MAC-VO

↓ 3 callersMethod__init__
(self, inplanes, planes, stride, downsample, pad, dilation)
Module/Network/TartanVOStereo/PSM/submodule.py:23
↓ 3 callersMethod__init__
(self, inp_dim, out_dim, kernel_size=3, stride=1, bn=False, relu=True)
Module/Network/TartanVOStereo/PSM/hourglass.py:6
↓ 3 callersMethod__init__
(self, config: SimpleNamespace)
Module/Frontend/Frontend.py:54
↓ 3 callersFunction_centerCropTo
(x: torch.Tensor, shape: int, dim: int)
Utility/Utils.py:67
↓ 3 callersFunction_cropToMultiple
(x: torch.Tensor, factor: int, dim: int)
Utility/Utils.py:39
↓ 3 callersFunction_padTo
(x: torch.Tensor, factor: int, dim: int, pad_value: float)
Utility/Utils.py:96
↓ 3 callersMethodapply
(self, func: T.Callable[[torch.Tensor,], torch.Tensor])
Module/Map/Graph.py:50
↓ 3 callersFunctioncreate_3x3_matrix
(matrix: list[list[torch.Tensor | float]], n_sample: int, device: torch.device)
Module/Covariance/Project2to3.py:426
↓ 3 callersMethodcropAndResize
(x: torch.Tensor, target_shape: tuple[int, int])
Module/Network/TartanVOStereo/StereoVO_Interface.py:159
↓ 3 callersMethoddeserialize
(cls, prefix: str, value: dict[str, np.ndarray])
Module/Map/Graph.py:130
↓ 3 callersFunctiondisparity_to_depth
(disp: torch.Tensor, bl: float, fx: float)
Module/Frontend/StereoDepth.py:271
↓ 3 callersMethodencoder
(self, im1, im2)
Module/Network/PWCNet/RAFTCov.py:20
↓ 3 callersMethodestimate_pair
Given two frames with imageL, imageR with shape of Bx3xHxW, return `output` of - [0] - IStereoDepth output of stereo frame from tim
Module/Frontend/Frontend.py:62
↓ 3 callersMethodfrom_evo
(cls, evo_traj: PoseTrajectory3D, frame_status: torch.Tensor | None=None)
Utility/Trajectory.py:135
↓ 3 callersMethodfrom_timed_SE3_numpy
(cls, file: Path, frame_status: torch.Tensor | None = None)
Utility/Trajectory.py:83
↓ 3 callersMethodget_leaves
(self)
Utility/Sandbox.py:119
↓ 3 callersMethodinference_2_match
(flow_12: torch.Tensor, cov_12: torch.Tensor)
Module/Frontend/Frontend.py:197
↓ 3 callersMethodlog_points
(rerun_path: str, position: torch.Tensor, color: torch.Tensor | None, cov_Tw: torch.Tensor | None, cov_mode: T
Utility/Visualize/Rerun_Visualize.py:110
↓ 3 callersFunctionnamespace_to_cfgnode
Design for Flowformer. Flowformer uses yacs.config.CfgNode as config container.
Utility/Config.py:86
↓ 3 callersFunctionplot_LostTrackAt
(ax: Axes, axis_0: int, axis_1: int, traj: PlotableTrajectory)
Utility/Plot/PlotTraj.py:86
↓ 3 callersFunctionplot_MotionROE_axes
(ax: Axes, ref_traj: PlotableMotions, est_traj: PlotableMotions, axis: int)
Utility/Plot/PlotTraj.py:62
↓ 3 callersFunctionplot_MotionRTE_axes
(ax: Axes, ref_traj: PlotableMotions, est_traj: PlotableMotions, axis: int)
Utility/Plot/PlotTraj.py:32
↓ 3 callersFunctionplot_Trajectory
(ax: Axes, axis_0: int, axis_1: int, traj: PlotableTrajectory)
Utility/Plot/PlotTraj.py:79
↓ 3 callersFunctionplot_experiment
(ax, disparity: float, gamma: float, bl: float, fx: float, num_sample: int)
Scripts/Statistic/Montecarlo_DisparityVariance.py:13
↓ 3 callersMethodplot_figure
(self, axis0_name: str="Axis 0", axis1_name: str="Axis 1", reduction: T.Literal["None", "Log"] = "None")
Utility/Extensions/GridRecorder.py:53
↓ 3 callersMethodplot_imatcher
Plot the IMatcher module output in various subplots. Show mask, covariance and flow as much as possible.
Utility/Visualize/Matplotlib_Visualize.py:114
↓ 3 callersMethodpredict
Estimate the pose of next frame given current frame, estimated depth and flow. NOTE: returned pose should be under global co
Module/MotionModel.py:25
↓ 3 callersMethodscale_targetflow
calculte GT flow in different scales
Module/Network/PWCNet/pwc/pwc_model_tartanvo.py:299
↓ 3 callersMethodset
(self, elem_idx: torch.Tensor, map_idx: torch.Tensor)
Module/Map/Graph.py:228
↓ 3 callersMethodset_autoremove
(self)
Utility/Sandbox.py:128
↓ 3 callersFunctionsimulate_cov_model
( ax: Axes, model: ICovariance2to3, bl: float, K: torch.Tensor, disparity: float, u: float, v: fl
Scripts/AdHoc/CovarianceModel.py:18
↓ 3 callersMethodterminate
You can define additional operations on terminate. For instance, smoothing trajectory / interpolate bad frames etc.
Odometry/Interface.py:85
↓ 3 callersMethodupsample_flow
(self, flow, mask)
Module/Network/PWCNet/pwc_cov/gru.py:40
↓ 2 callersFunctionAnalyzeRotation
(runs: list[tuple[PlotableMotions, PlotableMotions]], file_name: Path)
Utility/Plot/PlotFigure.py:84
↓ 2 callersFunctionAnalyzeTranslation
( runs: list[tuple[PlotableMotions, PlotableMotions]], file_name: Path )
Utility/Plot/PlotFigure.py:25
↓ 2 callersFunctionCovariance_2to3_full
Project the uncertainty on 2D image plane to 3D space (under camera coordinate) In * sigma_uu: torch.Tensor (N,) Variance of
Module/Covariance/Project2to3.py:377
↓ 2 callersFunctionMahalanobisDist
Argument x : torch.Tensor of shape N x F mu : torch.Tensor of shape N x F sigma : torch.Tensor of shape
Utility/Math.py:137
↓ 2 callersFunctionPlotTrajectory
(trajs: list[PlotableTrajectory], file_name: Path)
Utility/Plot/PlotFigure.py:138
↓ 2 callersMethod__getitem__
(self, index)
Module/Map/Graph.py:40
↓ 2 callersMethod__init__
(self, length: int)
DataLoader/SequenceBase.py:31
↓ 2 callersMethod__init__
(self, input_dim=128, hidden_dim=256)
Module/Network/FlowFormerCov/covhead.py:10
↓ 2 callersMethod__init__
(self, cfg)
Module/Network/PWCNet/pwc_cov/attention.py:91
↓ 2 callersMethod__push
Actual underlying circular buffer push algorithm.
Utility/Extensions/TensorExtension.py:231
↓ 2 callersMethod__write_scalar_batch
(self)
Utility/Extensions/TensorExtension.py:198
↓ 2 callersMethod_alloc_new_tensor
(self, shape, **kwargs)
Utility/Extensions/TensorExtension.py:80
↓ 2 callersMethod_get_emb
Gets a base embedding for one dimension with sin and cos intertwined
Module/Network/PWCNet/pwc_cov/attention.py:50
↓ 2 callersMethod_optimize
Given context and argument, construct the optimization problem, solve it and return the updated context and result.
Module/Optimization/Interface.py:111
↓ 2 callersMethodalign_time
(self, new_time: torch.Tensor)
Utility/Trajectory.py:172
↓ 2 callersMethodapply_mask
(self, cam_mask: np.ndarray)
DataLoader/Dataset/EuRoC.py:210
↓ 2 callersMethodapply_mask
(self, cam_mask: np.ndarray)
DataLoader/Dataset/VBR.py:42
↓ 2 callersMethodbuild_intrinsic
(intrinsic: list[float])
DataLoader/Dataset/EuRoC.py:177
↓ 2 callersFunctioncoords_grid
(batch, ht, wd)
Module/Network/PWCNet/pwc_cov/gru.py:8
↓ 2 callersFunctioncreate_2x2_matrix
(matrix: list[list[torch.Tensor | float]], n_sample: int, device: torch.device)
Module/Covariance/Project2to3.py:436
↓ 2 callersMethodcrop
(self, from_idx: int | None = None, to_idx: int | None = None)
Utility/Trajectory.py:193
↓ 2 callersFunctioncropToMultiple
(x: torch.Tensor, factor: int | list[int], dim: int | list[int])
Utility/Utils.py:37
↓ 2 callersMethodcrop_stereo
(data: StereoData, target_h: int, target_w: int)
DataLoader/Transform.py:109
↓ 2 callersFunctiondisparity_to_depth_cov
(disp: torch.Tensor, disp_cov: torch.Tensor, bl: float, fx: float)
Module/Frontend/StereoDepth.py:276
↓ 2 callersMethodelaborate_map
Given a sequence of frames, elaborate the trajectory (frame poses) and handle the 'need_interp' (i.e. lost track / skipped) frames.
Module/MapProcessor.py:20
↓ 2 callersMethodestimate
(self, frame: StereoData, kp: torch.Tensor, depth_est: IStereoDepth.Output, depth_cov: torch.Tensor | None, fl
Module/Covariance/Project2to3.py:52
↓ 2 callersMethodestimate_depth
Given stereo frames with imageL, imageR with shape of Bx3xHxW, return IStereoDepth `output` of stereo frame #### All outputs
Module/Frontend/Frontend.py:73
↓ 2 callersFunctionfinal_cov_loss
(preds: torch.Tensor, gt: torch.Tensor, cov_preds: list[torch.Tensor], flow_mask: torch.Te
Train/MatchingNet/loss.py:37
↓ 2 callersFunctionflow_loss
(gamma: float, preds: torch.Tensor, gt: torch.Tensor, mask: torch.Tensor)
Train/MatchingNet/loss.py:5
↓ 2 callersMethodforward_flow
(self, x0_flow, x0n_flow)
Module/Network/TartanVOStereo/StereoVO.py:26
↓ 2 callersMethodframe2Sample
Adapt the SourceDataFrame used in this project into the "sample" format used in TartanVO
Module/Network/TartanVOStereo/StereoVO_Interface.py:65
↓ 2 callersMethodframeRangeQuery
Retrieve IMU data in range of [start_frame, end_frame) Args: start_frame (int): start image frame to get IMU sequence
DataLoader/Dataset/EuRoC.py:329
↓ 2 callersFunctiongaussain_full_kernels
In: cov_2x2: torch.Tensor of shape Nx2x2, full 2D covaraiance matrices kernel_size: int, must be a positive odd number Ou
Utility/Math.py:44
↓ 2 callersMethodget_class
(cls: type[T], type: str)
Utility/Extensions/SubclassRegistry.py:29
↓ 2 callersMethodget_frame2match
(self, frame: FrameNode)
Module/Map/VisualMap.py:86
↓ 2 callersMethodget_match2point
(self, match: MatchObs)
Module/Map/VisualMap.py:89
↓ 2 callersFunctionget_sample_dimention
(sample)
Module/Network/TartanVOStereo/Utility.py:107
↓ 2 callersFunctioninitialize_flow
Flow is represented as difference between two means flow = mean1 - mean0
Module/Network/PWCNet/pwc_cov/gru.py:14
↓ 2 callersMethodinterpolate_vecN
(ev_time: torch.Tensor, time: torch.Tensor, value: torch.Tensor, N=3)
DataLoader/Dataset/EuRoC.py:364
↓ 2 callersMethodisKeyframe
(self, frame: StereoFrame)
Module/KeyframeSelector.py:23
↓ 2 callersMethodis_valid_config
(cls, config: SimpleNamespace | None)
Module/Covariance/Project2to3.py:57
↓ 2 callersMethodis_valid_config
(cls, config: SimpleNamespace | None)
Module/Frontend/Matching.py:117
↓ 2 callersFunctionloadTartanAirGT
(path: Path)
DataLoader/Dataset/TartanAir.py:452
↓ 2 callersMethodload_npy_format
(path: Path)
DataLoader/Dataset/TartanAir.py:270
↓ 2 callersMethodload_png_format
(path: Path)
DataLoader/Dataset/TartanAir.py:274
↓ 2 callersMethodlog_trajectory
(rerun_path: str, trajectory: pp.LieTensor | torch.Tensor, **kwargs)
Utility/Visualize/Rerun_Visualize.py:69
↓ 2 callersFunctionmake_intrinsics_layer
(w, h, fx, fy, ox, oy)
Module/Network/TartanVOStereo/Utility.py:6
↓ 2 callersFunctionmerge_matrices
(matrices)
Train/MatchingNet/train_flowformer.py:34
↓ 2 callersFunctionmove_dataclass_to_local
(obj: T_GraphInput)
Module/Optimization/Interface.py:27
↓ 2 callersMethodname
Assign a short name for the dataset class. By default will be the class name. Overwrite this function if you want to create a more
DataLoader/SequenceBase.py:20
↓ 2 callersFunctionplot_MotionROE
(ax: Axes, ref_traj: PlotableMotions, est_traj: PlotableMotions)
Utility/Plot/PlotTraj.py:51
↓ 2 callersFunctionplot_image
Expect Shape and range: Image - (H, W, 3) - Any range
Utility/Plot/PlotAxes.py:42
↓ 2 callersFunctionpoint2pixel_NED
(points: torch.Tensor, intrinsics: torch.Tensor)
Utility/Point.py:19
↓ 2 callersMethodpostprocess
(img, shape, isflow=False)
Module/Network/PWCNet/RAFTCov.py:183
↓ 2 callersMethodpreprocess
(img)
Module/Network/PWCNet/RAFTCov.py:171
↓ 2 callersMethodpush
push a batch of values into the CircularTensor. Will trigger a cache eviction for scalar writing buffer before pushing the value to e
Utility/Extensions/TensorExtension.py:223
↓ 2 callersMethodpush_keyframe
(self, frame: T_SensorFrame, est_pose: pp.LieTensor | torch.Tensor, need_interp: bool=False)
Odometry/MACVO.py:339
↓ 2 callersMethodpush_scalar
(self, value: int | float)
Utility/Extensions/TensorExtension.py:271
↓ 2 callersMethodrun
Core method for IVisualOdometry. This method handles the incoming frames and perform tracking/mapping internally.
Odometry/Interface.py:72
↓ 2 callersMethodrun
The main process that continuously running to manage different modules in MAC-VO. The multi-threading part will be managed in this fu
Odometry/MACVO.py:352
↓ 2 callersFunctionrun_frame
(cfg: Path, frame1: StereoFrame, frame2: StereoFrame)
Scripts/AdHoc/Optimization_Ablation.py:12
↓ 2 callersMethodscale_mask
threshold: deperated in tarranair, mask=0: valid_mask - True mask=1: CROSS_OCC - False mask=10:
Module/Network/PWCNet/pwc/pwc_model_tartanvo.py:313
↓ 2 callersMethodscale_stereo
(data: StereoData, scale_u: float, scale_v: float, interpolate: Literal["nearest", "bilinear"])
DataLoader/Transform.py:54
↓ 2 callersMethodunique_sync_mask
(stamps: np.ndarray, common_time: np.ndarray)
DataLoader/Dataset/VBR.py:148
↓ 2 callersMethodwrite_map
(self, global_map: VisualMap)
Module/Optimization/Interface.py:224
↓ 1 callersFunctionAnalyzeROE_cdf
(runs: list[tuple[PlotableMotions, PlotableMotions]], axis: Literal[0, 1, 2, None], file_name: Path)
Utility/Plot/PlotFigure.py:223
↓ 1 callersFunctionAnalyzeRTE_cdf
(runs: list[tuple[PlotableMotions, PlotableMotions]], axis: Literal[0, 1, 2] | None, file_name: Path)
Utility/Plot/PlotFigure.py:205
← previousnext →101–200 of 937, ranked by callers